Primary Outcomes
a. Hospital admission in ED studies: Five studies20,23,25,27,28 reported information for this outcome. In the pooled data of the four studies20,23,25,27 that reported absolute admission data, there was a tendency to favor budesonide for reducing the risk of hospital admission, but it did not reach a significant difference (RR= 0.89 [95% CI: 0.74 to 1.07], I2=0%, p=0.21) (Figure 3). This result was similar using random effect model (RR=0.92 [0.77-1.09), p=0.34). Razi et al.28 reported a significantly higher discharge rate in the budesonide than in the placebo group (survival analyses: log-rank=12.407, p<0.001).
In the subanalysis of studies that included only moderate to severe acute asthma episodes23,25, there was no significant difference in hospital admission between the budesonide and placebo groups (RR= 0.88 [0.68 to 1.14], I2=47%, p= 0.34). In the subanalysis of ICS doses (budesonide ≥ 2 mg vs. < 2mg), there was no significant difference in hospital admission between groups (RR=0.89 [0.71 to 1.11], I2=26%, p=0.29 vs. RR=0.89 [0.67 to 1.19], p=0.44).
b. Hospital LOS: Four studies20,21,26,27reported data for this outcome. Pooled data from two studies21,26 that reported hospital LOS in hours showed a significant reduction in LOS for those who received budesonide rather than controls (MD = -29.08 hours [-39.9 to -18.3]; I2=0%, p=<0.00001) (Figure 4). This result was similar using random effect model (data not shown). The other study20reported a survival analysis where children who received budesonide were released from the ED or discharged from the hospital significantly more rapidly than were those who received placebo (log rank test, p = 0.02); but Kassisse et al.27 reported no significant difference.